these. ness. In the first place, it is necessary, according to him, to distinguish between 'natural' and 'philosophical relation.' The latter is one of which the idea is acquired by the comparison of objects, as distinct from natural relation or 'the quality by which two ideas are connected together in the imagination, and the one naturally' (i.e. according to the principle of association) introduces the other' (p. 322). Of His acphilosophical relation-or, according to another form of ex- count of pression, of qualities by which the ideas of philosophical relation are produced '-seven kinds are enumerated; viz. ' resemblance, identity, relations of time and place, proportion in quantity and number, degrees in quality, contrariety, and causation' (ibid., and p. 372). Some of these do, some do not, apparently correspond to the qualities by which the mind is naturally conveyed from one idea to another;' or which, in other words, constitute the gentle force' that determines the order in which the imagination habitually puts together ideas. Freedom in the conjunction of ideas, indeed, is implied in the term 'imagination,' which is only thus differenced from memory;' but, as a matter of fact, it commonly only connects ideas which are related to each other in the way either of resemblance, or of contiguity in time and place, or of cause and effect. Other relations of the philosophical sort are the opposite of natural. Thus, distance will be allowed by philosophers to be a true relation, because we acquire an idea of it by the comparing of objects; but in a common way we say, "that nothing can be more distant than such or such things from each other; nothing can have less relation "' (ibid.). It corresponds to Locke's account of the sorts of agreement 207. Hume's classification of philosophical relations evidently serves the same purpose as Locke's, of the four sorts of agreement or disagreement between ideas,' in the perception of which knowledge consists; but there are some important discrepancies. Locke's second sort, which he between awkwardly describes as agreement or disagreement in the ideas. way of relation,' may fairly be taken to cover three of Hume's kinds; viz. relations of time and place, proportion in quantity or number, and degrees in any quality. About Locke's first sort, identity and diversity,' there is more difficulty. Under identity,' as was pointed out above, he includes the See above, paragraph 25 and the passages from Locke there referred to. Could relations which Hume distinguishes as identity proper' and 208. It is his method of dealing with these ideas of philoHume con- sophical relation that is specially characteristic of Hume. sistently admit idea Let us, then, consider how the notion of relation altogether is of relation affected by his reduction of the world of consciousness to ot all? 1 See above, paragraph 122. 2 Locke, Book IV. sec. iii. chap. xiv.; and above, paragraph 121 and 122. · Impressions and ideas. What is an impression? To this, as we have seen, the only direct answer given by him is that it is a feeling which must be more lively before it becomes. less so.1 For a further account of what is to be understood by it we must look to the passages where the governing terms of school-metaphysics' are, one after the other, shown to be unmeaning, because not taken from impressions. Thus, when the idea of substance is to be reduced to an unintelligible chimæra,' it is asked whether it be derived from the impressions of sensation or reflection? If it be conveyed to us by our senses, I ask, which of them, and after what manner? If it be perceived by the eyes, it must be a colour; if by the ears, a sound; if by the palate, a taste; and so of the other senses. But I believe none will assert that substance is either a colour, or a sound, or a taste. The idea of substance must therefore be derived from an impression of reflection, if it really exist. But the impressions of reflection resolve themselves into our passions and emotions' (p. 324). From the polemic against abstract ideas we learn further that the appearance of an object to the senses' is the same thing as an impression becoming present to the mind' (p. 327). That is to say, when we talk of an impression of an object, it is not to be understood that the feeling is determined by reference to anything other than itself: it is itself the object. To the same purpose, in the criticism of the notion of an external world, we are told that the senses are incapable of giving rise to the notion of the continued existence of their objects, after they no longer appear to the senses; for that is a contradiction in terms' (since the appearance is the object); and that they offer not their impressions as the images of something distinct, or independent, or external, because they convey to us nothing but a single perception, and never give us the least intimation of anything beyond' (p. 479). The distinction between impres sion of sensation and impression of reflection, then, cannot, any more than that between impression and idea, be regarded as either really or apparently a distinction between outer and inner. 'All impressions are internal and perishing existences' (p. 483); and, everything that enters the mind being in reality as the impression, 'tis impossible anything should to feeling appear different' (p. 480). See above, paragraphs 195 and 197. Only in regard to identity and causa tion that he sees any difficulty. These he fictions re- of ideas; 209. This amounts to a full acceptance of Berkeley's doctrine of sense; and the question necessarily arises-such being the impression, and all ideas being impressions grown weaker, can there be an idea of relation at all? Is it not open to the same challenge which Hume offers to those who talk of an idea of substance or of spirit? It is from some one impression that every real idea is derived.' What, then, is the one impression from which the idea of relation is derived? If it be perceived by the eyes, it must be a colour; if by the ears, a sound; if by the palate, a taste; and so of the other senses.' There remain our passions and emotions;' but what passion or emotion is a resemblance, or a proportion, or a relation of cause and effect? 210. Respect for Hume's thoroughness as a philosopher must be qualified by the observation that he does not attempt to meet this difficulty in its generality, but only as it affects. the relations of identity and causation. The truth seems to be that he wrote with Berkeley steadily before his mind; and it was Berkeley's treatment of these two relations in particular as not sensible but intelligible, and his assertion of a philosophic Theism on the strength of their mere intelligibility, that determined Hume, since it would have been an anachronism any longer to treat them as sensible, to dispose of them altogether. The condition of his doing so with success was that, however unwarrantably, he should treat the other relations as sensible. The language, which seems to express ideas of the two questionable relations, he has to <account for as the result of certain impressions of reflection, called 'propensities to feign,' which in their turn have to be accounted for as resulting from the natural relations of ideas according to the definition of these quoted above,' as the qualities by which one idea habitually introduces another.' Among these, as we saw, he included not only resemblance and contiguity in time or place, but cause and effect.' There is no relation,' he says, which produces a stronger connection in the fancy than this.' But in this, as in much of the language which gives the first two Parts their plausibility, he is taking advantage of received notions on the part of the reader, which it is the work of the rest of the book to set aside. In any sense, according to him, in which it differs Seo above, paragraph 206. from usual contiguity, the relation of cause and effect is itself reducible to a propensity to feign,' arising from the other natural relations; but when the reader is told of its producing 'a strong connection in the fancy,' he is not apt to think of it as itself nothing more than the product of such a connection. For the present, however, we have only to point out that Hume, when he co-ordinates it with the other natural relations, must be understood to do so provisionally. According to him it is derived, while they are primary. Upon them, then, rested the possibility of filling the gap between the i.e. from occurrence of single impressions, none determined by refer ence to anything other than itself,' and what we are pleased contiguity. to call our knowledge, with its fictions of mind and thing, of real and apparent, of necessary as distinct from usual connection. 211. We will begin with Resemblance. As to this, it will be said, it is an affectation of subtlety to question whether there can be an impression of it or no. The difficulty only arises from our regarding the perception of resemblance as different from, and subsequent to, the resembling sensations; whereas, in fact, the occurrence of two impressions of sense, such as (let us say) yellow and red, is itself the impression of their likeness and unlikeness. Hume himself, it may be further urged, at any rate in regard to resemblance, anticipates this solution of an imaginary difficulty by his important division of philosophical relations into two classes (p. 372)— such as depend entirely on the ideas which we compare together, and such as may be changed without any change in the ideas and by his inclusion of resemblance in the former class. resem blance and Is resemblance then an impres 212. Now we gladly admit the mistake of supposing that sensations undetermined by relation first occur, and that afterwards we become conscious of their relation in the sion? way of likeness or unlikeness. Apart from such relation, it is true, the sensations would be nothing. But this ad- \ mission involves an important qualification of the doctrine that impressions are single, and that the mind (according to Hume's awkward figure) is a bundle or collection of these,' succeeding each other in a perpetual flux or movement.' It implies that the single impression in its singleness is what it is through relation to another, which must there |